analyse-design_skill

This skill reverse-engineers a design system from code and screenshots to produce a developer-ready design reference for components.
  • Python

110

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill sammcj/agentic-coding --skill analyse-design

  • SKILL.md3.7 KB

Overview

This skill reverse-engineers an application's design system from its frontend codebase and any provided screenshots. It produces a practical design system reference developers can use to build consistent UI components. The output highlights colours, typography, spacing, component patterns, iconography, motion, and responsiveness with file-level citations when available.

How this skill works

I scan the project for theme/token files, global styles, component configs, layout and representative components using glob/grep heuristics. I extract explicit tokens (colors, fonts, spacings) and infer missing rules from CSS/JS values and screenshots. The result is a structured design reference with concrete tokens, usage notes, and places to fix inconsistencies.

When to use it

  • When you need a component library or style guide for an existing frontend codebase.
  • To audit visual consistency and find mismatched colours, spacing, or typography.
  • When onboarding designers or engineers to a legacy project without documentation.
  • Before implementing a redesign or migrating to a design-token system.
  • When you have screenshots and want the visual language codified into tokens.

Best practices

  • Provide theme/token files and representative screenshots for highest-fidelity results.
  • Point to primary layout and component files (shell, buttons, cards) to speed analysis.
  • Use consistent naming in tokens; I will call out one-off values to consolidate.
  • Include both light and dark screenshots if dark mode exists to capture variants.
  • Accept suggested token mappings and then run a follow-up pass after changes.

Example use cases

  • Generate a colour palette and semantic token mappings from tailwind.config.js and CSS variables.
  • Document typography scale and assign semantic roles (heading, body, caption) from fonts.css and components.
  • Audit button/input/card patterns and produce recommended border-radius, shadow, and state rules with source file links.
  • Identify responsive breakpoints from layout files and describe mobile adaptations.
  • Detect animation easing and durations from CSS/JS and summarize motion patterns for transitions.

FAQ

Provide theme/token files (tailwind.config, theme.ts/js, tokens.json), global styles (index.css), and representative component files (buttons, inputs, layout). Screenshots are highly recommended.

Can you infer values if tokens are missing?

Yes — I extract computed CSS values and sample pixels from screenshots to infer colours, spacing, and type scales, and I flag inferred items vs explicit tokens.

Will you modify the codebase?

No — I only read and analyze files and images. I produce a reference document and remediation suggestions for developers to apply.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational